SITCOM: Step-wise Triple-Consistent Diffusion Sampling for Inverse Problems

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Alkhouri, Ismail, Liang, Shijun, Huang, Cheng-Han, Dai, Jimmy, Qu, Qing, Ravishankar, Saiprasad, Wang, Rongrong
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910968159666176
author Alkhouri, Ismail
Liang, Shijun
Huang, Cheng-Han
Dai, Jimmy
Qu, Qing
Ravishankar, Saiprasad
Wang, Rongrong
author_facet Alkhouri, Ismail
Liang, Shijun
Huang, Cheng-Han
Dai, Jimmy
Qu, Qing
Ravishankar, Saiprasad
Wang, Rongrong
contents Diffusion models (DMs) are a class of generative models that allow sampling from a distribution learned over a training set. When applied to solving inverse problems, the reverse sampling steps are modified to approximately sample from a measurement-conditioned distribution. However, these modifications may be unsuitable for certain settings (e.g., presence of measurement noise) and non-linear tasks, as they often struggle to correct errors from earlier steps and generally require a large number of optimization and/or sampling steps. To address these challenges, we state three conditions for achieving measurement-consistent diffusion trajectories. Building on these conditions, we propose a new optimization-based sampling method that not only enforces standard data manifold measurement consistency and forward diffusion consistency, as seen in previous studies, but also incorporates our proposed step-wise and network-regularized backward diffusion consistency that maintains a diffusion trajectory by optimizing over the input of the pre-trained model at every sampling step. By enforcing these conditions (implicitly or explicitly), our sampler requires significantly fewer reverse steps. Therefore, we refer to our method as Step-wise Triple-Consistent Sampling (SITCOM). Compared to SOTA baselines, our experiments across several linear and non-linear tasks (with natural and medical images) demonstrate that SITCOM achieves competitive or superior results in terms of standard similarity metrics and run-time.
format Preprint
id arxiv_https___arxiv_org_abs_2410_04479
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SITCOM: Step-wise Triple-Consistent Diffusion Sampling for Inverse Problems
Alkhouri, Ismail
Liang, Shijun
Huang, Cheng-Han
Dai, Jimmy
Qu, Qing
Ravishankar, Saiprasad
Wang, Rongrong
Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
Diffusion models (DMs) are a class of generative models that allow sampling from a distribution learned over a training set. When applied to solving inverse problems, the reverse sampling steps are modified to approximately sample from a measurement-conditioned distribution. However, these modifications may be unsuitable for certain settings (e.g., presence of measurement noise) and non-linear tasks, as they often struggle to correct errors from earlier steps and generally require a large number of optimization and/or sampling steps. To address these challenges, we state three conditions for achieving measurement-consistent diffusion trajectories. Building on these conditions, we propose a new optimization-based sampling method that not only enforces standard data manifold measurement consistency and forward diffusion consistency, as seen in previous studies, but also incorporates our proposed step-wise and network-regularized backward diffusion consistency that maintains a diffusion trajectory by optimizing over the input of the pre-trained model at every sampling step. By enforcing these conditions (implicitly or explicitly), our sampler requires significantly fewer reverse steps. Therefore, we refer to our method as Step-wise Triple-Consistent Sampling (SITCOM). Compared to SOTA baselines, our experiments across several linear and non-linear tasks (with natural and medical images) demonstrate that SITCOM achieves competitive or superior results in terms of standard similarity metrics and run-time.
title SITCOM: Step-wise Triple-Consistent Diffusion Sampling for Inverse Problems
topic Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2410.04479